|Publication number||US7898280 B2|
|Application number||US 12/206,237|
|Publication date||Mar 1, 2011|
|Filing date||Sep 8, 2008|
|Priority date||Sep 8, 2008|
|Also published as||US20100060307|
|Publication number||12206237, 206237, US 7898280 B2, US 7898280B2, US-B2-7898280, US7898280 B2, US7898280B2|
|Original Assignee||Emil Kamieniecki|
|Export Citation||BiBTeX, EndNote, RefMan|
|Patent Citations (28), Non-Patent Citations (13), Referenced by (8), Classifications (10), Legal Events (1)|
|External Links: USPTO, USPTO Assignment, Espacenet|
This invention relates to electrical characterization of semiconductor materials.
The defining metric of what is competitive in the photovoltaics industry is the cost per kilowatt-hour. The success of a specific photovoltaic technology depends on its ability to reduce the manufacturing cost and increase the efficiency of the solar cells. In the integrated circuit (IC) industry, the cost of the substrate is frequently a negligible fraction of the total manufacturing cost. On the other hand, in the photovoltaics industry, the cost of the substrate material may ultimately determine the viability of the produce. The cost and the quality of substrates used for photovoltaic solar cells is critical in the manufacture of the cells.
More than 90% of photovoltaic solar cell manufacturers use silicon as a substrate material, and silicon wafers can account for as much as 45% of the total cost of photovoltaic modules. Hence, reducing the manufacturing cost of substrate materials without sacrificing quality is a potential opportunity for decreasing the cost of photovoltaics. Such an improvement, which is critical for both silicon and thin-film photovoltaics, should be achieved without a substantial reduction of the conversion efficiency of the photovoltaic cells.
Photovoltaic cells can be manufactured using electronic-grade silicon wafers; however, an increasingly common approach to reducing the cost of the photovoltaic substrate material is the use of lower quality (i.e., more impure) silicon sources and the replacement of single crystal silicon with multicrystalline silicon. Multicrystalline silicon can be fabricated by casting ingots and slicing the ingots into wafers. A growing number of manufacturers further reduce the cost of the substrate material by replacing sliced wafers with sheet or ribbon technology, which creates monocrystalline or multicrystalline silicon directly from a silicon melt. One such technology is the “String Ribbon Crystal Growth Technology” used by Evergreen Solar, Inc. Multicrystalline silicon suffers from a higher level of contamination due to the use of a less pure silicon source and also from a higher density of crystalline defects than single crystal (electronic-grade) silicon. Multicrystalline substrates are less expensive to manufacture, but photovoltaic cells made of multicrystalline silicon are also less efficient than cells made from single crystal silicon.
Transition metals are often a main culprit in degrading the efficiency of multicrystalline silicon solar cells. Multicrystalline photovoltaics also suffer from a large density of crystalline defects such as grain boundaries, as illustrated in
The use of thin films of photovoltaic material deposited directly onto rigid or flexible substrates offers additional cost savings. Thin film cells use around one-one hundredth the amount of semiconductor material used in single crystal or multicrystalline cells. Furthermore, in addition to silicon, the deposited photovoltaic material can be other inorganic semiconductors or even organic materials with optical properties better suited to the solar spectrum. In silicon, for instance, the theoretical upper limit of efficiency (i.e., the conversion of light to electricity) under normal full-sun illumination is 31%. This limit is largely a result of the 1.1 eV band gap of silicon, which corresponds to a photon wavelength of 1127 nm. In practice, systems using crystalline silicon as a photovoltaic material are only about 20-25% efficient. In contrast, a more efficient photovoltaic cell is a three-junction indium gallium phosphide/indium gallium arsenide/ germanium cell, which increases the conversion efficiency to 37%. Fundamentally different devices, however, could have a theoretical efficiency as high as 87%. However, because non-silicon photovoltaic material is often based on compound semiconductors (e.g., copper indium gallium diselenide, or CIGS), the concentration of defects in such thin films can be even higher than the concentration in silicon.
Improving manufacturing processes and increasing manufacturing yields in photovoltaic fabrication requires reliable information regarding critical electrical properties of the substrate materials. One of the most critical parameters, the minority carrier lifetime τ, defines the quality of the photovoltaic substrate material and the efficient of the solar cells. The minority carrier lifetime, a measure of how long carriers (i.e., electrons and holes) remain before recombining, is highly sensitive to the presence of defects detrimental to device performance. A related parameter, the minority carrier diffusion length, is the average distance a carrier moves from its point of generation before it recombines. The longer the lifetime, the longer the diffusion length, and the higher the probability that optically generated carriers will reach the collection point (i.e., the p-n junction) and produce current in an external circuit. A shorter lifetime translates to a lower conversion efficiency of the solar cell and a poorer economic viability of the product.
The dominant approaches to measurement of the carrier lifetime in the photovoltaic industry are photoconductive decay (PCD) methods. In this technique, electron-hole pairs are created by an optical excitation that changes the conductivity of the sample. The recombination of excess carriers is monitored by measuring the decay of the conductivity after the illumination is turned off. Optical excitation with a pulse having a photon energy that exceeds the energy gap of the semiconductor generates excess electron-hole pairs, increasing the conductivity of the semiconductor (i.e., photoconductance). Relaxation of this photoconductance to the initial equilibrium state is detected using either radio frequency techniques (RF-PCD) or microwave reflectance techniques (μ-PCD). Both RF-PCD and μ-PCD measure an ‘effective total lifetime.’ In this approach, separation of the bulk lifetime from the surface recombination requires surface passivation, such as immersion of silicon in HF, iodine passivation, or corona charging. Alternatively, the requirement for surface passivation can be reduced by measuring slabs of material at least ten times thicker than standard photovoltaic substrates and using deeply penetrating illumination (i.e., longer wavelength). In μ-PCD, a microwave beam is directed to the sample and the reflected microwave power, which is proportional to the conductance of the sample, is measured. μ-PCD systems, unlike RF-PCD systems, exhibit a non-linear dependence on the injection level, and thus meaningful μ-PCD measurements are limited to a narrow range of excess carrier concentrations.
Another approach to measuring the minority carrier lifetime is the surface photovoltage (SPV) method. SPV measures the minority carrier diffusion length, which is directly related to the minority carrier lifetime and which can be used interchangeably with the lifetime in material characterization. An advantage of SPV over PCD is the ability of SPV to distinguish between bulk and surface carrier recombination. However, SPV requires a sample thickness larger than the carrier diffusion length which, for typical materials used in commercial solar cells (carrier lifetime ˜2-5 μs), exceeds the standard thickness of silicon photovoltaic wafers (around 100-200 μm). Additionally, in the SPV method, uncontrolled surface recombination at the wafer backside may strongly interfere with carrier recombination in the bulk, leading to misleading data. Furthermore, extended imperfections such as grain boundaries and dislocations can also interfere with SPV measurements, reducing their reliability.
Differentiation between recombination processes occurring in the bulk of the device and at the interfaces can be accomplished by identifying the dependence of lifetime on the geometry of the device being tested or on the excitation wavelength. Determining the underlying physics of the recombination process can be accomplished by analyzing the dependence of the lifetime on the level of photo-excitation and temperature. Such an approach, while cumbersome and time consuming, is quite effective in the R&D environment; however, it is not suitable for in-line monitoring of the manufacturing process.
The quasi-steady-state photoconductance (QSSPC) method measures the dependence of the lifetime (multiplied by the carrier mobility) on the level of photo-excitation from a quasi steady-state value of the photoconductance. In QSSPC, as in the PCD method, separation of the bulk lifetime from surface recombination requires surface passivation unless performed on slabs of material at least ten times thicker than standard photovoltaic substrates.
Separation of the effects of surface or extended crystallographic imperfections from the effect of bulk contaminants and bulk point defects is critical for process optimization and control of the fabrication of photovoltaics. The ability to maintain a low concentration of electrically active metallic impurities and crystalline defects is essential for achieving a high yield in silicon manufacturing and thus for reducing the cost of photovoltaic devices.
This invention relates generally to a method and a system for characterizing electronic properties of a semiconductor sample.
In one aspect, a method for characterizing electronic properties of a semiconductor sample includes illuminating the surface of the semiconductor sample with a pulse of light, measuring a photoconductance decay in the semiconductor sample after the cessation of the first pulse of light, and analyzing the photoconductance decay. The electronic properties include properties associated with at least one of the bulk of the semiconductor sample and the surface of the semiconductor sample. The pulse of light has a predetermined duration and a photon energy higher than an energy gap of the semiconductor. The analyzing step determines a first component of the photoconductance decay substantially associated with point imperfections in the semiconductor sample and at least one second component of the photoconductance decay substantially associated with extended imperfections in the semiconductor sample.
In embodiments of this aspect of the invention, illuminating the surface of the semiconductor sample and measuring the photoconductance decay is performed at a constant distance from the semiconductor sample.
If the bulk of the semiconductor sample includes only point imperfections and the only extended imperfection is the surface of the semiconductor sample embodiments may include one or more of the following. The electronic properties of the bulk of the semiconductor sample are determined by analyzing the first component of the photoconductance decay and the electronic properties of the surface are determined by analyzing the second component of the photoconductance decay.
If the bulk of the semiconductor sample includes at least one of point imperfections and extended imperfections and the surface of the semiconductor sample is passivated, electronic properties of the bulk of the semiconductor sample associated with point imperfections are determined by analyzing the first component of the photoconductance decay and electronic properties of the bulk of the semiconductor sample associated with extended imperfections are determined by analyzing the second component of the photoconductance decay. The method includes determining a ratio of a first volume of the semiconductor sample in which point imperfections dominate carrier recombination to a second volume of the semiconductor sample in which extended imperfections dominate carrier recombination.
If the electronic properties of the bulk of the semiconductor sample include electronic properties associated with point imperfections and electronic properties associated with extended imperfections in the bulk and the surface of the semiconductor sample is substantially not passivated, the electronic properties associated with point imperfections are determined by analyzing the first component of the photoconductance decay. The electronic properties associated with extended imperfections in the bulk and the electronic properties of the surface of the semiconductor sample are determined by analyzing the second component of the photoconductance decay.
Illuminating the surface of the semiconductor sample includes illuminating the surface of the semiconductor sample with a first pulse of light having a first wavelength and a first intensity. Measuring a photoconductance decay includes measuring a first photoconductance decay. Analyzing the photoconductance decay includes analyzing the first photoconductance decay. The method includes illuminating the surface of the semiconductor sample with a second pulse of light having a second wavelength and a second intensity, measuring a second photoconductance decay in the semiconductor sample after the cessation of the second pulse of light, analyzing the second photoconductance decay to determine a third component of the photoconductance decay substantially associated with extended imperfections in the semiconductor sample, and distinguishing between the electronic properties associated with extended imperfections in the bulk and electronic properties of the surface of the semiconductor sample on the basis of the second component of the photoconductance decay and the third component of the photoconductance decay. At least one of the second wavelength and the second intensity is different from the first wavelength and the first intensity. The method includes determining a ratio of a first volume of the semiconductor sample in which point imperfections dominate charge carrier recombination to a second volume of the bulk of the semiconductor sample in which extended imperfections dominate carrier recombination. The pulse of light includes at least two wavelengths.
The electronic properties of the surface of the semiconductor sample comprise at least one of a surface charge density, a density of surface traps, and a surface potential barrier. The density of surface traps is used to monitor a surface cleaning or etching processes. The electronic properties of the bulk of the semiconductor sample comprise at least one of a probability per unit time that a majority carrier will be captured by an available point imperfection or extended imperfection, an effective volume of extended imperfections in the bulk of the semiconductor sample, a bulk carrier lifetime, and a subsurface doping concentration. The subsurface doping concentration is determined and used to calibrate an apparatus performing the measurement. The subsurface doping concentration is determined and used to characterize a subsurface region of the semiconductor sample. The subsurface doping concentration is determined and used to characterize a thin film.
The pulse of light has a duration sufficient to achieve steady-state photoconductance in the sample. An interval between consecutive pulses of light is sufficient to allow the sample to substantially reach thermal equilibrium. Measuring the photoconductance decay in the semiconductor sample includes performing the measurement at a distance from the surface of the sample. The distance is between 1 μm and 300 μm, e.g., between 25 μm and 100 μm. The first component of the photoconductance decay is approximated by an exponential function. The second component of the photoconductance decay is approximated by a logarithmic function. The method includes generating inversion conditions at the surface of the semiconductor sample prior to illuminating the surface of the semiconductor sample. The inversion conditions are generated using a chemical treatment or corona charge.
Illuminating the surface of the semiconductor sample can further include varying the intensity of the pulse of light, varying the photon energy, or both.
In another aspect, a method for characterizing electronic properties of a semiconductor sample includes illuminating a surface of the semiconductor sample for a duration of time sufficient to establish steady-state photoconductance in the semiconductor sample, measuring a non-linear dependence of a photoconductance in the semiconductor sample on the intensity of the illumination, and analyzing the non-linear dependence of the photoconductance on the intensity of the illumination to determine electronic properties of the semiconductor sample. In an embodiment, the electronic properties of the sample include at least one of a surface charge density, a density of surface electronic traps, a bulk carrier lifetime, or a subsurface doping concentration. In another embodiment, measuring the photoconductance decay in the semiconductor sample includes performing the measurement at a distance from the surface of the sample.
In a further aspect, a system for characterizing electronic properties of a semiconductor sample includes a light source, an inductive probe, an RF source, and a controller. The light source is positioned to illuminate the semiconductor sample with a pulse of light having a predetermined duration and a photon energy higher than an energy gap of the semiconductor sample. The inductive probe is spaced from the semiconductor sample and positioned to provide an RF signal to the semiconductor sample. The inductive probe is further configured to detect a photoconductance signal from the semiconductor sample in response to the pulse of light. The RF source provides the RF signal having a frequency to the inductive probe. The controller tunes the frequency of the RF signal provided to the inductive probe on the basis of the semiconductor sample
In embodiments of this aspect of the invention, the controller can automatically tune the frequency of the RF signal. The system can further include an RF bridge electrically connected between the RF source and inductive probe. The RF bridge has a resonance branch including the inductive probe and a reference branch including a mechanism for varying the magnitude and phase of the reference branch.
The controller can be configured to control the mechanism for varying the magnitude and phase of the reference branch as well as to control the light source.
The inductive probe includes an optical fiber coupled to the light source and a detector coil and can further include an electromagnetic shield surrounding the optical fiber and the detector coil. The inductive probe is spaced from the semiconductor sample at a fixed distance.
The controller can be configured to perform the steps of illuminating the surface of the semiconductor sample with a pulse of light having a predetermined duration and a photon energy higher than an energy gap of the semiconductor; measuring a photoconductance decay in the semiconductor sample after the cessation of the first pulse of light; and analyzing the photoconductance decay to determine a first component of the photoconductance decay substantially associated with point imperfections in the semiconductor sample and at least one second component of the photoconductance decay substantially associated with extended imperfections in the semiconductor sample.
In this embodiment, an RF bridge can be electrically connected between the RF source and inductive probe, the RF bridge having a resonance branch including the inductive probe and a reference branch including a mechanism for varying the magnitude and phase of the reference branch. The RF bridge can be further configured to be balanced before measuring the photoconductance decay in the semiconductor sample.
The use of a method or system for characterizing electronic properties of a semiconductor sample has a number of advantages. In particular, because the effects of various material imperfections are controlled by different process parameters, effective process control requires the ability to monitor the electrical effect of each type of imperfection separately. A non-contact, non-destructive system and method as described above can be used for rapid, in-line monitoring and characterization of the carrier lifetime in semiconductor samples. Bulk and surface effects are separated in a single measurement step without requiring surface passivation. The effect of bulk point imperfections on the carrier recombination can be distinguished from the effect of other critical types of material imperfections such as extended imperfections. Such a metrology system and method enables better process optimization and in-line process control of photovoltaic fabrication, improving yields of functioning devices and reducing the cost of photovoltaic solar cells.
Semiconductor Characterization System
A system and a method for measuring the photoconductance of a semiconductor sample are discussed below. Referring for example to
Probe 200 contains an optical fiber 204, such as a 1 mm fused silica fiber or an acrylic fiber having an inner diameter of 1 mm, for illuminating sample 202. A detection coil 206 in the form of a surface coil is coplanar with the end of optical fiber 204 closest to semiconductor sample 202. Detection coil 206 is composed of multiple windings 208 (for instance, approximately three windings, or more specifically approximately 2¾ windings). The axis of optical fiber 204 is aligned with the center of windings 208; optical fiber 204 is held in place relative to windings by a fiber optic collar (not shown). Surrounding optical fiber 204 and detection coil 206 is a cylindrical copper shield 210. An end of one of the windings 208 is electrically connected through connections 212 to copper shield 210 and to a shield of a semi-rigid coaxial cable 207. In another embodiment, copper shield 210 is not grounded. Any space between optical fiber 204, windings 208, and copper shield 210 is filled with an insulating epoxy. The bottom of the insulating epoxy is optically polished to achieve a planar surface at the base of probe 200. The base of probe 200 is adjusted using an optical mount to be parallel to the surface of sample 202.
Probe 200 is not in contact with sample 202; instead, a constant separation between sample 202 and the base of probe 200 is automatically maintained by using a conventional electromechanical actuator or stage (not shown) controlled by an optical or capacitive distance-measuring sensor. For instance, the separation is between about 1 μm and 300 μm, or between about 25 μm and 100 μm.
In one embodiment, the detection coil is wound around an optical collimator and the optical fiber is coupled to the collimator. Use of a collimator reduces the dependence of the light intensity of the separation between the sample and the detection coil.
Referring again to
Controller 304, which includes one or more computers 306 with programmable processors, controls the output of light source 302 and is coupled to light source 302 via a driver 308. A pulse generator 310 in controller 304 controls the intensity of the light emitted by light source 302 and the duration and repetition frequency of the illumination pulses, which may be rectangular pulses. Controller 304 also generally includes an analog-to-digital convertor (not shown) and a digital signal processor 366.
Sample 202 is also subjected to radio frequency (RF) electromagnetic radiation, preferably ultrahigh frequency (UHF) electromagnetic radiation. A frequency synthesizer 312, such as an oscillator, acts as an RF source and generates an RF signal 314, which passes through a power splitter 316. Frequency synthesizer 312 is controlled by controller 304. Power splitter 316 sends a first portion 314 a of RF signal 314 to an RF bridge 318 and a second portion 314 b through an amplifier 320 to a phase detector 322 (see below). Frequency synthesizer 312 may be, for instance, model PTS500 from Programmed Test Sources, Inc., Littleton, Mass., capable of providing a signal in a frequency range from 1 MHz to 500 MHz. RF bridge 318 is constructed around a 180-degree hybrid 324, which is a passive RF coupling device that includes four ports. An example of a suitable hybrid is model HC-W500-MS available from Universal Microwave Components Corporation (UMCC, Alexandria, Va.). When a signal is applied at a sum port (Σ port) 326 of hybrid 324, that signal is equally divided in phase and amplitude and appears at both a 0-degree port 328 and a 180-degree port 330. Any signal reflected from 0-degree port 328 appears in phase at a delta port (Δ port) 332; any signal reflected from 180-degree port 330 appears 180 degrees out of phase at Δ port 332.
Signal 314 a provided by frequency synthesizer 312 is applied to Σ port 326 of hybrid 324. One-half of signal 314 a is sent into a resonance branch 336 (see below); any power reflected from resonance branch 336 appears in phase at Δ port 332. The other half of signal 314 a is sent through 0-degree port 328 to a reference branch 338. Reference branch 338 functions to balance RF bridge 318, as described below in conjunction with
Resonance branch 336 contains probe 200. In one embodiment, coaxial cable 207 of probe 200 is connected to 180-degree port 330 of hybrid 324 through an RF transformer 324, such as two transformers of model TC4-11, manufactured by Mini-Circuits, Brooklyn, N.Y., connected in series. This configuration ensures that the input impedance of hybrid 324, for instance 50 Ohms, is matched. In another embodiment, the diameter of detection coil 206 and the number of windings 208 are selected such that the input impedance of hybrid 324 is closely matched. In this case, coaxial cable 207 is connected directly to 180-degree port 330 of hybrid 324 without the need for a transformer.
The signal returning from resonance branch 336 exits hybrid 324 from Δ port 332 as a signal 346. Signal 346 passes through a first attenuator 348, a first amplifier 350, a low-pass filter 352, and reaches a power splitter 354. Power splitter 354 sends a first portion 346 a of signal 346 to an RF power meter 356 and a second portion 346 b of signal 346 through a second amplifier 358 and a second attenuator 360 to an RF phase detector 322. RF phase detector 322 demodulates signal 346 a into two output DC signals 362 a and 362 b representing quadrature components of incoming signal 346 a. Output signals 362 a and 362 b enter controller 304 and pass through low-pass filters 364 a and 364 b, respectively, which remove high frequency components from the signals. Signals 362 a and 362 b then pass through the analog-to-digital converter (not shown) and are processed by digital signal processor (DSP) 366 and computer 306. This process is performed multiple times, digitized, averaged, and, after subtraction of offsets, the results are combined to determine a magnitude of the signal. Use of the signal magnitude reduces the phase noise of the system and reduces the sensitivity to phase adjustment in reference branch 338 of RF bridge 318.
Referring again to
After RF bridge 318 is balanced, a photoconductance measurement is performed on sample 202 (426). A change in photoconductance in sample 202 affects the impedance between sample 202 and probe 200 and results in a loss of balance of RF bridge 318 and a corresponding change in signal 346 exiting Δ port 332. The resulting RF signal is submitted to RF phase detector 322, where it is split into output signals 362 a and 362 b. These output signals are filtered in controller 304 by low-pass filters 364 a and 364 b. The signal is repetitively measured a predetermined number of times. Each time, the signal is digitized for processing in DSP 366 and subject to averaging, subtraction of offsets, and analysis in computer 306.
Referring again to
After validation of the measurement parameters, characterization system 300 performs measurement, filtering, and averaging of a predetermined number of pulse sequences for each component of the signal at phase detector 322 (608). The results of the averaging step are tested for consistency (610) using criteria similar to those of the measurement validation process, with appropriate standard deviation limits. Measurement offsets, determined by the magnitude of the signal during time period t1, are subtracted from the results of the averaging process (612). The magnitude of the signal is calculated (614) and the results are analyzed (616). Use of the signal magnitude reduces phase noise of the system and reduces sensitivity to the phase adjustment performed in reference branch 338 and resonance branch 336. The signal analysis is performed by controller 304. In one embodiment, RF bridge is fine balanced before each measurement pulse sequence.
Semiconductor Characterization Methods
It is possible to characterize a semiconductor sample by analyzing the dependence of a steady-state photoconductance on the intensity of an illumination light or by analyzing the time dependence of a photoconductance decay. For ease of explanation, the methods below are described with reference to semiconductor characterization system 300. However, it is to be understood that these methods are not limited to use only with characterization system 300; any system or apparatus that enables performance of the methods below may be used. Furthermore, the discussions and formulas herein are directed to an n-type semiconductor sample. However, the discussions and formulas could easily be converted to apply to a p-type semiconductor as well. The term ‘photoconductance’ is used herein to describe the sheet photoconductance of a sample. The unit for sheet photoconductance used in presenting formulas and measurement results herein is the Siemen (S).
The contribution of various types of imperfections to carrier recombination is extracted from either the steady-state photoconductance or the photoconductance decay. For instance, imperfections having different recombination characteristics or imperfections associated with different regions of a sample (e.g., a bulk of the sample or a surface of the sample) can be distinguished. The discussion herein divides imperfections into two classes based on their recombination characteristics: point imperfections and extended imperfections.
Point imperfections, or zero-dimensional (0-D) imperfections, include single-atom impurities, point defects in the crystal lattice of a sample, and small clusters including a few impurity atoms or lattice defect atoms. The photoconductance decay (that is, the decay of the photoconductance after illumination is stopped) of point imperfections is characterized by an exponential decay; the steady-state photoconductance exhibits a linear dependence on the intensity of the illumination or on the associated optical carrier generation rate. The photoconductance decay associated with point imperfections is given by
Δσ0-D=Δσ0-D(0)exp(−t/τ 0-D) (1)
where Δσ0-D(0) is the steady-state photoconductance or, more generally, the initial photoconductance at time t=0, immediately at the cessation of the illumination pulse; and τ0-D is the effective minority carrier recombination lifetime due to recombination at point imperfections. If various types of point imperfection recombination centers are present in a sample, the effective recombination lifetime is a combination of the lifetimes associated with each type of point imperfection.
The steady-state photoconductance associated with point imperfections is given by
Δσ0-D(0)=q(μe+μh)G 0-Dτ0-D (2)
where q is the magnitude of the elementary charge, μe and μh are the mobilities of the electron and the hole, respectively, and G0-D is the generation rate in the vicinity of a point imperfection.
Extended material imperfections, also known as multi-dimensional (M-D) imperfections, include, for example, one-dimensional (1-D) defects such as edge dislocations, two-dimensional (2-D) defects such as surfaces and grain boundaries, and three-dimensional (3-D) imperfections such as large (i.e., consisting of a large number of atoms) clusters of impurities or lattice defects. Under certain conditions, extended imperfections trap majority carriers (electrons for n-type semiconductors), thus becoming charged. The charge produces an electrostatic potential barrier Ψd at the imperfection, limiting further capture of majority carriers. In some cases, such as at a surface, a charge arises due to built-in charge, such as a fixed oxide charge at an interface between silicon and silicon dioxide. Charged extended imperfections induce a Schottky-type depletion space charge region in the vicinity of the imperfection.
The control of the carrier capture rate by the height of the electrostatic potential barrier Ψd leads to a logarithmic decay of the photoconductance associated with extended imperfections. If the carrier generation rate is not too high and thus does not substantially reduce the electrostatic potential barrier, most of the photogenerated minority carriers (holes) are captured by the imperfections and free electrons dominate the photoconductance. In this case, the photoconductance decay associated with extended imperfections is given by
where time t begins at the cessation of the illumination pulse, nt0 is the density of electrons trapped at the imperfection in thermal equilibrium (i.e., in dark conditions), and τ0 and τe are time constants. Time constant τ0 is given by the formula
and time constant τe is given by the formula
where ft0 is the probability that an imperfection state is occupied by an electron in thermal equilibrium, Ψd0 is the thermal equilibrium value of the electrostatic potential barrier, ce is the probability per unit time that an electron will be captured by an available imperfection site, and n0 is the free electron concentration at thermal equilibrium.
For one-dimensional defects such as edge dislocations, the parameter Ξ is given as
for two-dimensional defects such as surfaces or grain boundaries, Ξ is given by the formula
where Qd0 is the charge density at the defect at thermal equilibrium, εs is the dielectric permittivity of the semiconductor material, and ND is the concentration of donors in the semiconductor. It should be noted that Ψd0 and Qd0, and hence Ξ, are negative for n-type semiconductors; for p-type semiconductors these parameters are positive, but the polarity of Eqs. (4)-(6) is also reversed so that the resulting photoconductance and all time constants remain positive.
At high carrier generation rates, the steady-state photoconductance in the vicinity of extended imperfections is described by Eq. (4) for t=0, and can also be written as
displaying a logarithmic (i.e., non-linear) dependence on the intensity of the illumination pulse.
At a sufficiently low carrier generation rate in the vicinity of an imperfection, the change in the potential barrier, ΔΨd, becomes much lower than the thermal energy kT, i.e.,
and the photoconductance decay in the vicinity of an extended imperfection becomes an exponential process described by the formula
ΔσM-D L=ΔσM-D L(0)exp(−t/τ e), (11)
where Δσ M-D L(0) is the photoconductance at t=0. The time constant τe plays effectively the same role played by the minority carrier lifetime in the photoconductance decay for point imperfections.
Steady-state photoconductance in the vicinity of extended imperfections at low carrier generation rates is described by the formula
ΔσM-D L(0)=qμ e G M-Dτe. (12)
This formula has the same form as the formula for steady-state photoconductance associated with recombination at point imperfections; however, instead of bipolar photoconductance, it considers only photoconductance associated with majority carriers.
Steady-state photoconductance associated with extended imperfections, given by Eqs. (9) and (12), considers only the contribution of electrons (majority carriers). Depending on measurement parameters, such as the wavelength of the illumination pulse and the material properties of the sample, the steady-state photoconductance could be distorted by the presence of excess free holes (minority carriers) not trapped in the space charge region of the imperfection. Similarly, the photoconductance decay, given by Eqs. (4) and (11), considers only the contribution of electrons. Depending on the measurement configuration, there could still be remnant holes outside of the imperfection space charge region at the first moment after the illumination pulse is stopped. However, these remnant holes would rapidly drift toward the space charge region and, a short time after the beginning of the relaxation, the decay process would be governed essentially by the rate of electron capture.
G 0 =c e n 0 N t exp(qΨ d0 /kT). (13)
The linear dependence shown for comparison represents light intensity dependence of the steady-state photoconductance associated with point imperfections. It demonstrates that linear photoconductance would dominate at high enough intensities while photoconductance associated with extended imperfections would dominate at low intensities.
5 × 1014
−2 × 1010
−2 × 1010
−2 × 1010
−2 × 1010
−4 × 1010
1 × 1011
1 × 1011
5 × 1010
1 × 1011
1 × 1011
9.65 × 10−7
9.65 × 10−7
9.65 × 10−7
9.65 × 10−7
9.65 × 10−7
Curves A and B both have a logarithmic portion 1100 with the same slope which, for extended imperfections, is determined based on Equation (4) to be
thus indicating that curves A and B correspond to samples having the same ND/Qd0 (doping concentration over defect charge density) ratio. Both curves are also characterized by the same time constant τe which, for two-dimensional defects, can be written following Equation (6) as
indicating that the same electronic properties of the defects are responsible for carrier recombination. The only difference between curves A and B is the generation rate G. The difference in generation rate gives rise to different time constants τ0 which, for two-dimensional defects, can be written following Equation (5) as
Comparison of curves A and B shows that the initial decay of the photoconductance, shown in a region 1102 of curve A and a region 1104 of curve B, strongly depends through τ0 on the generation rate G. This initial portion of the decay is followed by the logarithmic region 700 described by Equation (4) and subsequently by an exponential decay (not shown) described by Equation (11).
Curves A and B demonstrate that, under high illumination conditions, the photoconductance decay does not follow the exponential decay pathway conventionally considered in photoconductance decay measurements. Furthermore, it shows that the slope of a decay curve does not necessarily characterize the decay. In particular, the initial slope of the decay (e.g., regions 1102 and 1104 of curves A and B) is determined by steady-state conditions and depends on the carrier generation rate. The recombination rate, or capture rate, is determined by the time at which a linear (in a semi-log plot) extension of the logarithmic portion of the decay crosses zero photoconductance.
Curves A and B were calculated using a density of defect traps, Nt, of 1011 cm−2 and a probability of electron capture, ce, of 9.65×10−7 cm 3/s, the latter of which corresponds to a typical electron capture cross section for defect states present at an interface between silicon and silicon dioxide. To illustrate the effect of the electronic properties of two-dimensional defect states on the photoconductance decay, curve C was calculated using the same generation rate as curve A but with the density of defect traps reduced by a factor of two. The photoconductance decay is also affected by the defect charge density, Qd0, and the doping concentration, ND, in the region surrounding the imperfections. Curve D corresponds to a sample having the same generation rate as curve A but with the defect charge density increased by a factor of two; curve E has the doping concentration decreased by a factor of two.
The above-described dependence of the photoconductance on material properties can be used as a basis for materials characterization. In particular, in addition to or instead of considering exponential decay and linear dependence of the photoconductance on the illumination intensity (known as a linear or low-intensity approximation and corresponding to the low-excitation case discussed above), the non-linear photoconductance behavior characterized by non-exponential decay and non-linear dependence of the photoconductance on the illumination intensity is used to extract material properties from a photoconductance measurement. The following sections describe three embodiments demonstrating specific applications of the above method. The data analysis in the embodiments below may be performed automatically by controller 304 shown in
In one embodiment, electronic properties of a surface of a semiconductor sample are characterized by photoconductance measurements of the sample. Before measurement, the sample is prepared for analysis by forming a depletion or inversion layer at the surface. Some samples may already possess a surface charge sufficient to induce a depletion or inversion layer due to previous processing; other samples are treated with a chemical treatment or by deposition of a corona charge. In silicon, chemical treatment with SC-1 or HF, depending on the sample type (i.e., n- or p-type), is conventionally used for such treatment. In one embodiment, surface preparation is performed before the sample is placed in the characterization system. Alternatively, the surface may be charged using corona discharge by the characterization system during measurement. In one embodiment, the corona charging conditions required for achieving surface inversion are established empirically by gradually depositing ionic corona charge and measuring either τe or the steady-state photoconductance; a stable value of τe or of the steady-state photoconductance indicates that inversion conditions have been established at the surface. Since τe depends exponentially on the height of the surface potential barrier, monitoring τe results in a more accurate determination of inversion than monitoring of the steady-state photoconductance.
The wavelength of the illumination pulse is selected to be sufficiently short so as to generate electron-hole pairs predominantly in the vicinity of the surface, so that all minority carriers are swept into the surface space charge region. For silicon, a diode laser emitting around 400 nm would provide sufficient illumination; however, even at longer wavelengths, such as 660 nm, the light absorption in the vicinity of the surface is sufficient to ensure dominance of the surface region. The intensity of the illumination and the length of the illumination pulse are selected to provide an excitation level (i.e., a total density of excess carriers) sufficient to change the surface potential barrier by an amount greater than the thermal energy (i.e., the situation of Eq. (3)) but low enough to ensure that the surface potential barrier remains substantially higher than the thermal energy. Since the photoconductance associated with the surface as an extended imperfection depends logarithmically on the intensity of illumination (see Eq. (9)) and the photoconductance associated with point imperfections in the bulk depends linearly on the illumination intensity (see Eq. (2)), reducing the carrier injection level (i.e., decreasing the illumination intensity) results in a decreased contribution of the bulk to the measured photoconductance (see
Under such conditions, using Eqs. (4)-(6) and the expression for Ξ2-D, the photoconductance decay can be expressed in a more explicit form given by the formula
and where it is assumed that, at room temperature, n0=ND.
The steady-state photoconductance described by Eq. (9) can be rewritten as
for the above conditions, where Qs0 and Ψs0 are the surface charge and the height of the surface potential barrier, respectively, at thermal equilibrium.
The measurement conditions are pre-established empirically by verifying that the photoconductance decay includes a logarithmic decay segment. The length of the pulse is selected to achieve steady-state photoconductance during illumination. In other embodiments, the pulse length is shorter than the duration required to achieve steady-state conditions.
Using known material parameters, such as εs, μe, and ND (for n-type semiconductors), the slope of the logarithmic component is used to determine the thermal equilibrium value of the surface charge density Qd0 (1210). Alternatively, the surface charge density Qd0 is determined from the slope of the dependence of the steady-state photoconductance on the generation rate (ln G in Eq. (20)). Using known material parameters, the independently determined carrier excitation rate G, and the surface charge density Qd0 determined in step 1210, τ0 is determined from Eq. (17) (1212). From t0 and τ0, τe is then determined (1214). For sufficiently high illumination intensities, τ0 becomes negligible and t0 gives the value of τe directly. From these calculated parameters, the value of ceNt exp(qΨd0/kT) can be determined (1216). Alternatively, ceNt exp(qΨd0/kT) is calculated from the dependence of the steady-state photoconductance on the generation rate G. Following Eq. (20), a linear extension (in semi-log scale) of the logarithmic component of the photoconductance decay to the point at which this extension crosses zero photoconductance (see
c e N t exp(qΨ d0 /kT)=G 0 /n 0. (22)
The parameter ceNt exp(qΨd0/kT) can be simplified by pre-charging the surface into inversion conditions using chemical treatment or corona charging. Since the value of Ψd0 under inversion is known, ceNt can be directly determined.
System Calibration and Determination of the Subsurface Doping Concentration
The above procedure for surface characterization can be used for system calibration and determination of a subsurface doping concentration. Determination of the subsurface doping concentration using surface photovoltage methods is commonly used by silicon wafer manufacturers. This approach requires inversion conditions (a maximum depletion layer width) induced by chemical treatment or corona charging. The method presented herein enables this approach to be extended to thinner wafers and thin films. It also offers a more complete characterization of the sample, including determination of the bulk recombination lifetime, described below, in a rapid and non-destructive manner.
Referring again to
where Qd0 is the surface charge density and ni is an intrinsic carrier density. Equation (23) applies to n-type semiconductors having surface inversion; for p-type semiconductors, the polarity of the surface charge is positive and the right side of Equation (23) does not have a negative sign. Since the slope of the logarithmic component of the photoconductance decay depends only on basic material parameters and on the subsurface doping concentration (that is, doping within the surface space charge region), the slope can be used either to calibrate the system (1220), if doping concentration is known, or to determine the doping concentration for a previously calibrated system (1222).
Cleaning Process Monitoring
In one embodiment, the method for surface characterization described above is used to monitor surface cleaning and etching processes. This Defect-Specific Photoconductance (DSPC) technique is based on a non-linear RF-photoconductance measurement that monitors such processes.
The density of surface traps (i.e., surface states) was calculated for the two samples depicted in
TABLE 2 Sample of FIG. 13 Sample of FIG. 14 Doping Concentration, ND [cm−3] 1.06 8.3 Surface Charge Density, Qs0 [q/cm2] −9 −7.87 Time Constant τe [ms] 2.0 4.1 ce 5.06 2.47 Nt [1/cm2] (for ce = 9.65 5.24 2.56
Separation of Bulk and Surface Properties
In another embodiment, both bulk and surface properties of a semiconductor sample are characterized simultaneously without surface passivation.
The wavelength of the illumination pulse is selected to be sufficiently energetic so as to generate electron-hole pairs beyond the surface space charge region and into the bulk of the sample. The intensity of the illumination and the length of the illumination pulse are selected to provide an excitation level (i.e., a total density of excess carriers) sufficient to change the surface potential barrier by an amount greater than the thermal energy but low enough to ensure that the surface potential barrier remains substantially higher than the thermal energy. Since the photoconductance associated with the surface depends logarithmically on the intensity of illumination (see Eq. (9)) and the photoconductance associated with point imperfections in the bulk depends linearly on the illumination intensity (see Eq. (2)), adjustment of the carrier injection level (i.e., the illumination intensity) allows the bulk and surface contributions to the measured photoconductance to be balanced. Since surface effects dominate at low illumination intensities (see
The measurement conditions are pre-established empirically by verifying that the photoconductance decay includes a logarithmic component and an exponential component. The length of the pulse is selected to achieve steady-state photoconductance during illumination. In other embodiments, the pulse is shorter than the duration required to achieve steady-state photoconductance.
Under these conditions, the overall photoconductance is equal to the sum of the photoconductance associated with surface recombination (described by Eq. (17)) and that associated with the bulk. Assuming that bulk recombination is dominated by point imperfections (characterized by Eqs. (1) and (2)), the overall photoconductance is given as
Gsc is the generation rate in the surface region, Gb is the generation in the bulk (both having units of 1/cm2s), and τb is the effective recombination lifetime in the bulk. The overall generation rate is given by
G=G sc +G b. (27)
The ratio of Gsc to Gb depends on the wavelength of the illumination pulse and on material properties such as doping concentration, surface properties, and bulk diffusion properties. In one embodiment, surface and bulk effects are distinguished by illuminating the sample with two consecutive pulses, each having a different wavelength, and analyzing the subsequent photoconductance decay resulting from each pulse. Alternatively, more than one wavelength can be contained in a single illumination pulse.
Referring again to
Contribution of Point and Extended Imperfections
The photoconductance measurements described above are useful for the characterization of semiconductor samples and for the monitoring and identification of various types of material imperfections. The method is particularly suitable for characterization of multi-crystalline semiconductors. In solar grade multicrystalline semiconductors such as silicon, the carrier lifetime is degraded due to both grain boundaries and chemical contaminants such as metals. These two sources of degradation are caused by different factors. The size of crystallites in the sample, which defines the density of centers associated with grain boundaries, depends to a large extent on parameters of the manufacturing process, whereas the concentration and type of contaminants depends on both processing and the purity of the starting material (e.g., polycrystalline silicon). Therefore, separating the contribution of these two types of imperfections is critical for photovoltaic manufacturing.
In one embodiment, the contribution to recombination properties and carrier lifetimes by point imperfections and extended imperfections can be distinguished. The procedure for distinguishing the contribution from each type of imperfection is closely related to the procedures described above. In this embodiment, the wavelength of the illumination pulse is selected to generate electron-hole pairs throughout the bulk of the sample. That is, the energy of the illumination pulse must be greater than the band gap of the semiconductor but sufficiently close to the main absorption edge of the sample so as to allow penetration of the light deep into the bulk of the sample. The intensity of the illumination and the length of the illumination pulse are selected to provide an excitation level (i.e., a total density of excess carriers) sufficient to change the surface potential barrier by an amount greater than the thermal energy but low enough to ensure that the surface potential barrier remains substantially higher than the thermal energy. Since the photoconductance associated with extended imperfections depends logarithmically on the intensity of illumination (see Eq. (9)) and the photoconductance associated with point imperfections depends linearly on the illumination intensity (see Eq. (2)), adjustment of the carrier injection level (i.e., the illumination intensity) allows contributions of point and extended imperfections to the measured photoconductance to be balanced.
The measurement conditions are pre-established empirically by verifying that the photoconductance decay includes a logarithmic component and an exponential component. The length of the pulse is selected to achieve steady-state photoconductance during illumination. In other embodiments, the pulse is shorter than the duration required to achieve steady-state photoconductance.
If necessary, the sample is prepared for analysis by passivation of a surface of the sample. Some samples may already be passivated due to previous processing. Passivation of, for example, silicon could be done by chemical treatment with SC-1 or HF depending on semiconductor conductivity type (p or n) or iodine passivation, which are conventionally used for such treatments. In one embodiment, surface preparation is done before the specimen is placed in the sample holder, e.g., by chemical treatment or by depositing a corona charge. Alternatively the specimen may be charged by the semiconductor characterization system during measurement.
Once the surface is passivated, the procedure described in
In this alternative embodiment, distinguishing contributions of the extended and point imperfections in the bulk to the recombination properties of semiconductor specimen is accomplished without surface passivation. In effect, this embodiment allows distinguishing contributions of both bulk point and bulk extended imperfections as well as the surface. Unlike the previous alternative, this embodiment requires the use of two series of pulses of different photon energy and different intensity. If two sources are used, their illumination could be combined into one optical fiber and delivered to the specimen through the fiber with the probe of
In this embodiment, characterization of the electronic properties of the specimen combines two series of measurements performed using excitation pulses of different wavelength and intensity. Wavelength (such as 830 nm laser diode) of the excitation pulses in one series of measurements is selected to generate electron-hole pairs through an entire depth of the specimen, generating carriers both in the vicinity of the surface as well as in the bulk. In this case carrier recombination occurs at the surface as well as at the extended and at the point imperfections in the bulk. The decay of the photoconductance following these pulses is illustrated in
In another embodiment, different types of imperfections are distinguished by analyzing the dependence of the steady-state photoconductance on the intensity of the illumination light.
Other embodiments not explicitly described herein are also within the scope of the claims. For instance, if generation rates for a certain type of sample are not low enough to reveal the logarithmic component of the photoconductance decay or not high enough to suppress an initial deviation from logarithmic decay (that is, if τ0 is too high), measurement could be performed at two or more generation rates and time constants could be determined by numerical iteration. Additionally, electronic properties of a sample could be obtained by analyzing the time dependent rise in photoconductance when an illumination pulse first strikes the sample. The wavelength dependence of photoconductance also provides data for distinguishing between different types of imperfections in a sample. The methods described herein are of particular interest in solar cell manufacturing and, more generally, in any semiconductor manufacturing process utilizing thin films, such as large panel displays. Furthermore, these methods can be used for real-time in-line testing of semiconductor materials.
It is to be understood that the foregoing description is intended to illustrate and not to limit the scope of the invention, which is defined by the scope of the appended claims. Other embodiments are within the scope of the following claims.
|Cited Patent||Filing date||Publication date||Applicant||Title|
|US2859407||Mar 27, 1956||Nov 4, 1958||Sylvania Electric Prod||Method and device for measuring semiconductor parameters|
|US3939415||Jul 24, 1974||Feb 17, 1976||Hitachi, Ltd.||Method of and device for measuring life time of carriers of semiconductor|
|US4000458||Aug 21, 1975||Dec 28, 1976||Bell Telephone Laboratories, Incorporated||Method for the noncontacting measurement of the electrical conductivity of a lamella|
|US4286215||May 18, 1979||Aug 25, 1981||Bell Telephone Laboratories, Incorporated||Method and apparatus for the contactless monitoring carrier lifetime in semiconductor materials|
|US4891584||Mar 21, 1988||Jan 2, 1990||Semitest, Inc.||Apparatus for making surface photovoltage measurements of a semiconductor|
|US4949034||Mar 13, 1989||Aug 14, 1990||Mitsubishi Kinzoku Kabushiki Kaisha||Method for contactless evaluation of characteristics of semiconductor wafers and devices|
|US5047713||Nov 14, 1990||Sep 10, 1991||Semitex Co., Ltd.||Method and apparatus for measuring a deep impurity level of a semiconductor crystal|
|US5406214||Dec 16, 1991||Apr 11, 1995||Semilab Felvezeto Fizikai Lab, Rt||Method and apparatus for measuring minority carrier lifetime in semiconductor materials|
|US5451886||Jun 16, 1993||Sep 19, 1995||School Judicial Person Ikutoku Gakuen||Method of evaluating lifetime of semiconductor material and apparatus for the same|
|US5477158||Oct 4, 1993||Dec 19, 1995||Loral Vought Systems Corporation||Compact noncontact excess carrier lifetime characterization apparatus|
|US5661408||Mar 1, 1995||Aug 26, 1997||Qc Solutions, Inc.||Real-time in-line testing of semiconductor wafers|
|US5929652||Sep 2, 1997||Jul 27, 1999||Midwest Research Institute||Apparatus for measuring minority carrier lifetimes in semiconductor materials|
|US6037797||Jul 11, 1997||Mar 14, 2000||Semiconductor Diagnostics, Inc.||Measurement of the interface trap charge in an oxide semiconductor layer interface|
|US6104206 *||Aug 5, 1997||Aug 15, 2000||Verkuil; Roger L.||Product wafer junction leakage measurement using corona and a kelvin probe|
|US6275060||Apr 1, 1999||Aug 14, 2001||Midwest Research Institute||Apparatus and method for measuring minority carrier lifetimes in semiconductor materials|
|US6315574||Jan 20, 2000||Nov 13, 2001||Qc Solutions, Inc.||Method for real-time in-line testing of semiconductor wafers|
|US6369603||Jul 14, 2000||Apr 9, 2002||Midwest Research Institute||Radio frequency coupling apparatus and method for measuring minority carrier lifetimes in semiconductor materials|
|US6538462 *||Nov 30, 1999||Mar 25, 2003||Semiconductor Diagnostics, Inc.||Method for measuring stress induced leakage current and gate dielectric integrity using corona discharge|
|US6909302||Aug 17, 2001||Jun 21, 2005||Qc Solutions, Inc.||Real-time in-line testing of semiconductor wafers|
|US6911350||Mar 28, 2003||Jun 28, 2005||Qc Solutions, Inc.||Real-time in-line testing of semiconductor wafers|
|US6967490||Mar 28, 2003||Nov 22, 2005||Qc Solutions, Inc.||Real-time in-line testing of semiconductor wafers|
|US7019513 *||Jan 19, 2005||Mar 28, 2006||Vladimir Faifer||Non-contact method and apparatus for measurement of sheet resistance and leakage current of p-n junctions|
|US7084661 *||May 18, 2001||Aug 1, 2006||Sensorchem International Corporation||Scanning kelvin microprobe system and process for analyzing a surface|
|US7439757 *||May 30, 2006||Oct 21, 2008||Lg Display Co., Ltd.||Apparatus and method for inspecting liquid crystal display|
|US7504838 *||May 2, 2006||Mar 17, 2009||Advanced Micro Devices, Inc.||Methods of determining characteristics of doped regions on device wafers, and system for accomplishing same|
|US7656172 *||Jan 18, 2006||Feb 2, 2010||Cascade Microtech, Inc.||System for testing semiconductors|
|US7663385 *||Dec 15, 2003||Feb 16, 2010||Nanometrics Incorporated||Apparatus and method for electrical characterization by selecting and adjusting the light for a target depth of a semiconductor|
|US20070273400 *||Dec 15, 2003||Nov 29, 2007||Emil Kamieniecki||Apparatus and Method for Electrical Characterization of Semiconductors|
|1||Adria Nieswand, "Thin is in", R&D Magazine, Dec. 2006.|
|2||D.T. Stevenson, and R.J. Keyes, "Measurement of Carrier Lifetimes in Germanium and Silicon," J. Appl. Phys. 26, pp. 190-195 (1955).|
|3||Dieter K. Schroder, "Semiconductor Material and Device Characterization," John Wiley & Sons, Inc., New York 1990, Chapter 7.|
|4||E. Yablonowitch, and T.J. Gmitter, "A Contactless Minority Carrier Lifetime Probe of Heterostructures, Surfaces, Interfaces, and Bulk Wafers", Solid-State Electron. 35, Mar. 1987, pp. 261-267.|
|5||G.L. Miller, D.A.H. Robinson, and J.D. Wiley, "Contactless measurement of semiconductor conductivity by radio frequency free carrier power absorption", Rev. Sci. Instrum. vol. 47(7), 1976, pp. 799-805.|
|6||H. M'saad, J. Michel, J.J. Lappe, and L.C. Kimerling, "Electronic Passivation of Silicon Surfaces by Halogens", J. Electron. Mat. 23, May 1994, pp. 487-491.|
|7||H.K. Henish and J. Zucker, "Contactless Method for the Estimation of Resistivity and Lifetime of Semiconductors", Rev. Sci. Instrum. 27, pp. 409-410 (Jun. 1956).|
|8||Industry Strategy Symposium (ISS 2008), SEMI. half Moon Bay, CA [Alexander E. Braun, Semiconductor International, Jan. 29, 2008].|
|9||J.C. Brice and P. Moore, Contactless Resistivity Meter for Semiconductors, J. Sci. Instrum. 38, p. 307 (Jul. 1961).|
|10||P.J. Olshefski, "Contactless Method for Measuring Resistivity of Silicon," Semic. Prod. 4, pp. 34-36 (Dec 1961).|
|11||S.M. Sze, "Physics of Semiconductor Devices", John Wiley & Sons, Inc., New York 1981, Chapter 7.|
|12||Sinton, R.A. and Cuevas, A., "Contactless determination of current-voltage characteristics and minority-carrier lifetimes in semiconductors from quasi-steady-state photoconductance data", Appl. Phys. Lett. vol. 69(17), 1996, pp. 2510-2512.|
|13||T. Figielski, "Theory of Carrier Recombination at Dislocations in Germanium", phys. stat. sol. 6, pp. 429-440 (1964) and "Dislocations as Traps for Holes in Germanium", phys. stat. sol. 9, pp. 555-566 (1965).|
|Citing Patent||Filing date||Publication date||Applicant||Title|
|US8364428 *||Nov 17, 2009||Jan 29, 2013||Imec||Junction-photovoltage method and apparatus for contactless determination of sheet resistance and leakage current of semiconductor|
|US8896338 *||Jun 14, 2012||Nov 25, 2014||Emil Kamieniecki||Electrical characterization of semiconductor materials|
|US8912799 *||Nov 9, 2012||Dec 16, 2014||Semiconductor Physics Laboratory Co., Ltd.||Accurate measurement of excess carrier lifetime using carrier decay method|
|US9002677 *||Jul 21, 2014||Apr 7, 2015||Raja Technologies||System and method of semiconductor characterization|
|US9140742||Feb 13, 2013||Sep 22, 2015||Samsung Display Co., Ltd.||Method of measuring a silicon thin film, method of detecting defects in a silicon thin film, and silicon thin film defect detection device|
|US20100153033 *||Nov 17, 2009||Jun 17, 2010||Imec||Junction-photovoltage method and apparatus for contactless determination of sheet resistance and leakage current of semiconductor|
|US20130169283 *||Nov 9, 2012||Jul 4, 2013||Semiconductor Physics Laboratory Co., Ltd.||Accurate measurement of excess carrier lifetime using carrier decay method|
|US20130257472 *||Jun 14, 2012||Oct 3, 2013||Emil Kamieniecki||Electrical characterization of semiconductor materials|
|Cooperative Classification||H01L22/14, G01R31/308, G01R31/2648, Y02E10/50, H01L31/186, Y02P70/521|
|European Classification||H01L31/18G, G01R31/26P|